Sparse regression problems, where it is usually assumed that there are many variables and that the effects of a large subset of variables are negligible, have become increasingly important. This paper describes the construction of hierarchical pri